telmisartan inhibits cell proliferation by blocking ...telmisartan inhibits cell proliferation by...

Telmisartan Inhibits Cell Proliferation by BlockingNuclear Translocation of ProHB-EGF C-TerminalFragment in Colon Cancer CellsKeiji Ozeki1, Satoshi Tanida1*, Chie Morimoto2, Yoshimasa Inoue3, Tsutomu Mizoshita1,

Hironobu Tsukamoto1, Takaya Shimura1, Hiromi Kataoka1, Takeshi Kamiya1, Eiji Nishiwaki3,

Hiroshi Ishiguro4, Shigeki Higashiyama5,6*, Takashi Joh1

1Department of Gastroenterology and Metabolism, Nagoya City University Graduate School of Medical Sciences, Nagoya, Aichi, Japan, 2Department of Living Science

Nutrition Course, Matsuyama Shinonome Junior College, Matsuyama, Ehime, Japan, 3Carna Biosciences Incorporation, Kobe, Japan, 4Department of Medical Technology

School of Health Sciences, Gifu University of Medical Science, Seki, Gifu, Japan, 5Department of Cell Growth and Tumor Regulation, Proteo-Medicine Research Center,

Ehime University, Toon, Ehime, Japan, 6Department of Biochemistry and Molecular Genetics, Ehime University Graduate School of Medicine, Ehime University, Toon,

Ehime, Japan

Abstract

Background & Aims: Current treatment target toward advanced colorectal cancers is mainly focused on the epidermalgrowth factor receptor (EGFR) signaling, but its additive effects with chemotherapy are still limited. A disintegrin andmetalloproteinase (ADAM) cleaves the proheparin-binding epidermal growth factor like growth factor (proHB-EGF). Andsoluble HB-EGF activates EGFR. In parallel, the carboxy-terminal fragment of proHB-EGF (HB-EGF-CTF) translocates into theinner nuclear membrane, and subsequently exerts on the regulation of cell proliferation by binding nuclear promyelocyticleukemia zinc finger (PLZF) protein, a transcriptional repressor, thereby causing its nuclear export. We hypothesized that theinhibition of HB-EGF-CTF nuclear translocation may be a new strategy in preventing cell proliferation.

Methods: 12-O-tetradecanoylphorbor-13-acetate (TPA) was treated to activate ADAM. Nine-thousand chemical compoundswere screened for their efficacies in blocking the binding of HB-EGF-CTF to promyelocytic leukemia zinc finger (PLZF) withAlphascreen system. The obtained candidates were then used to block the binding of HB-EGF-CTF to PLZF in colon cancercells, HT29 and HCT116. Cell proliferation was investigated with a growth curve assay. The intracellular localization, andassociation between HB-EGF-CTF and PLZF, was assessed with immunofluorescent staining, and immunoprecipitation andWestern blotting, respectively. The effects of obtained candidates on EGFR phosphorylation and on nuclear translocation ofHB-EGF-CTF and export of PLZF during the angiotensin II type1 receptor (AT1R) knockdown were also investigated.

Results: Telmisartan and candesartan were found to be potential candidates. Telmisartan inhibited TPA-induced cellproliferation stronger than candesartan. Telmisartan, but not candesartan blocked the nuclear translocation of HB-EGF-CTF,and binding of HB-EGF-CTF to PLZF, during TPA stimulation. Both telmisartan and candesartan did not inhibit TPA-inducedEGFR phosphorylation, and telmisartan, but not candesartan, inhibited TPA-induced nuclear translocation of HB-EGF-CTFafter knockdown of AT1R.

Conclusions: The inhibition of HB-EGF-CTF nuclear translocation with telmisartan may be a novel strategy in preventing cellproliferation.

Citation: Ozeki K, Tanida S, Morimoto C, Inoue Y, Mizoshita T, et al. (2013) Telmisartan Inhibits Cell Proliferation by Blocking Nuclear Translocation of ProHB-EGFC-Terminal Fragment in Colon Cancer Cells. PLoS ONE 8(2): e56770. doi:10.1371/journal.pone.0056770

Editor: Jian-Xin Gao, Shanghai Jiao Tong University School of Medicine, China

Received August 1, 2012; Accepted January 15, 2013; Published February 22, 2013

Copyright: � 2013 Ozeki et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This work was supported by a Grant-in Aid for scientific research C (KAKENHI 22590704) from the Japan Society for the Promotion of Science (JSPS). Thefunders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have the following Competing interest: Yoshimasa Inoue and Eiji Nishiwaki are employed by commercial company CarnaBiosciences Incorporation. There are no further patents, products in development or marketed products to declare. This does not alter the authors’ adherence toall the PLOS ONE policies on sharing data and materials, as detailed online in the guide for authors. All the authors declare that no financial conflict of interestexists in relation to the content of the article.

* E-mail: [email protected] (ST); [email protected] (SH)

Introduction

Colorectal cancer is the leading cause of death from gastroin-

testinal malignancy [1]. Currently therapeutic treatments toward

advanced colorectal cancers are mainly focused on the design of

therapeutic agents that is targeted to the epidermal growth factor

receptor (EGFR), the monoclonal blocking antibodies such as

cetuximab and panitumumab [2,3]. The additive effects of these

therapeutic antibodies plus combination chemotherapy on ad-

vanced colorectal cancer are obtained, but limited because the

KRAS mutation abolished these effects [3,4].

PLOS ONE | www.plosone.org 1 February 2013 | Volume 8 | Issue 2 | e56770

N-terminal fragments of EGFR ligands including heparin-

binding epidermal growth factor like growth factor (HB-EGF)

which is considered to be involved in mainly cell proliferation in

colon cancer cells bind to EGFR and induce its phosphorylation

[5,6,7,8,9]. G-protein coupled receptor (GPCR) agonists (e.g.

interleukin-8) and 12-O-tetradecanoylphorbol-13-acetate (TPA)

induced EGFR transactivation through a disintegrin and metallo-

proteinase (ADAM) - cleaved proHB-EGF [10,11]. The carboxy -

terminal fragments of proHB-EGF (HB-EGF-CTF) produced by

ectodomain shedding translocate into the inner nuclear mem-

brane, subsequently exert on the regulation of cell proliferation by

binding nuclear promyelocytic leukemia zinc finger (PLZF)

protein, a transcriptional repressor, thereby causing its nuclear

export [12,13]. Therefore, inhibition of nuclear translocation

signaling of HB-EGF-CTF is considered to be a new strategy

against the development of colorectal cancer. However, there is no

evidence of therapeutic agents that specifically block nuclear

translocation of HB-EGF-CTF and binding of that to PLZF.

Thus, the aims of the present study were to: i) screen for potent

chemical compounds that are capable of inhibiting the interactions

between HB-EGF-CTF and PLZF via GST-pull down assay,

surface plasmon resonance (SPR) spectroscopy and Alphascreen

technology, ii) validate the inhibitory effects of these potent

compounds on cell proliferation, and iii) certify the importance of

inhibiting the nuclear translocation of HB-EGF-CTF in colon

cancer cell proliferation.

Materials and Methods

MaterialsTPA was purchased from Cell Signaling Technology (Berverly,

MA, USA). Anti-EGFR polyclonal antibody, anti-phosphotyrosine

monoclonal antibody (4G10) and normal IgG were purchased

from Millipore (Billerica, MA, USA). The EGFR tyrosine kinase

inhibitor AG1478 and anti-PLZF antibody were purchased from

Calbiochem (LA Jolla, CA, USA). Candesartan was obtained from

Takeda Pharmaceutical Company Limited (Osaka, Japan) and

telmisartan was a kind gifted from Boehringer Ingelheim

(Ingelheim, Germany). Olmesartan and losartan were purchased

from Daiichi Sankyo Company limited (Tokyo, Japan) and Merck

Sharp & Dohme (Whitehouse Station, NJ, USA), respectively. The

anti-HB-EGF-CTF polyclonal antibody [12] and metalloprotei-

nase inhibitor, KB-R7785 [14], were prepared. The monoclonal

anti-FLAG M2 antibody and the peroxisome proliferator-activat-

ed receptor (PPAR) c antagonist, GW9662 [15,16], were

purchased from Sigma-Aldrich (St. Louis, MO, USA).

Plasmid PreparationA plasmid encoding FLAG-tagged PLZF was generated by

subcloning the FLAG sequence and human PLZF cDNA into

pcDNA3.1 (Invitrogen, Carlsbad, CA, USA). Plasmids for the

recombinant expression of FLAG-tagged PLZF derivatives, GST-

fused EGFR ligands-CTF, and CFP (cyan fluorescein protein)-

tagged PLZF were prepared [12]. Zinc fingers (ZnF) 5-8, and,

Figure 1. Dual signaling pathways of EGFR phosphorylation and HB-EGF C-terminal fragment nuclear translocation during cellproliferation. Quoted from [19] and modified. TPA induces an ADAM-mediated cleavage of proHB-EGF, and results in the ectodomain shedding ofits N-terminal fragment and generation of an intracellular C-terminal fragment (CTF). The soluble HB-EGF binds to the EGFR and induces a rapidtransient phosphorylation of EGFR. This phosphorylation results in the transcription of various genes. Meanwhile the HB-EGF-CTF is translocated intothe nucleus, where it subsequently induces the nuclear export of PLZF. This results in the progression of cell cycle. The potent inhibitor blocks thenuclear translocation of HB-EGF-CTF. P indicates phosphorylation. Abbreviations: EGFR; epidermal growth factor receptor, TPA; 12-O-tetradecanoylphorbol-13-acetate, PKCd; protein kinase Cd, ADAM; a disintegrin and metalloproteinase, HB-EGF; heparin-binding EGF-like growthfactor, CTF; C-terminal fragment, MAPK; mitogen-activated protein kinase, PLZF; promyelocytic leukemia zinc finger.doi:10.1371/journal.pone.0056770.g001

Telmisartan Inhibits Cell Growth via HB-EGF-CTF


Figure 2. Binding of EGFR ligand-CTFs to ZnF5-8 region of PLZF. (A) Schema of FLAG-tagged full length PLZF consisting of FLAG, BTB,Center, and nine ZnFs. (B) HT1080 cells stably expressing pro-HB-EGF, pro-TGF-a, pro-AR and pro-EPR were transiently transfected with an expressionvector encoding FLAG-tagged PLZF. PLZF protein expression of cell lysates (lane1), as well as, cells treated with 100 nM TPA for 1 h, and probed withanti-FLAG antibodies following immunoprecipitation with anti-EGFR ligands-CTF antibodies (lane2) and an anti-normal rabbit IgG (lane3). (C) GSTpull-down assay. Cell lysates containing FLAG-tagged PLZF derivatives were incubated with GST (lane 2), GST-HB-EGF-CTF (lane 3), GST-TGF-a-CTF(lane4), GST-AR-CTF (lane 5), and GST-EPR-CTF (lane 6) beads for 2 h, and bound proteins were detected by immunoblotting with an anti-FLAGantibody. (D) Schema of FLAG-tagged PLZF derivatives. The binding properties of PLZF derivatives to GST-fused- HB-EGF-CTF, TGF-a-CTF, AR-CTF andEPR-CTF GST in a pull-down assay, as summarized in the right lanes of each structure. Binding properties are based on the estimation of bandintensity with are relative to the control band and indicated by++(.50%),+(50–10%), and 2 (,10%).doi:10.1371/journal.pone.0056770.g002



Figure 3. The interaction of EGFR-ligand-CTFs and PLZF, and high-throughput screening for inhibitors blocking the interaction. (A)Direct interaction between EGFR-ligand-CTF and PLZF (GST-tagged-ZnF5-8) with the SPR system. The recombinant biotin-EGFR- ligand-CTFs wereindividually immobilized, by SA sensor chips. GST-Zn5-8 with concentrations ranging from 0.03 to 1.0 mM was then injected with running buffer at25uC and a flow rate of 30 mL/min for 2 min. (B) The expression vector encoding CFP-PLZF was co-transfected into wild-type HT1080 cells and HT1080cells stably expressing either proHB-EGF, proTGF-a, proAR or proEPR. The cells were cultured for 24 h and then pretreated in serum-free medium withKB-R7785. Cells were then treated in the serum-free conditioned medium with TPA, and the subcellular localization of the CFP fusion protein wasobserved. To determine the percentage of cells (mean6SD) demonstrating nuclear localization of CFP-PLZF, cells were counted in at least two



EGFR-CTF were cloned into pGEX6P-1(GE Healthcare, Little

Chalfont, U.K.).

Cell CultureHuman colon cancer cell lines, HT29 and HCT116 (American

Type Culture Collection, Rockville, MD, USA), were cultured

with McCoy’s5A (Sigma), CaCo2 (American Type Culture

Collection) were cultured with Dulbecco’s modified Eagle’s

medium (DMEM) with 20% fetal bovine serum (FBS). And

SW480 (American Type Culture Collection) was cultured DMEM

with 10% FBS. HT1080 cells and primary human keratinocytes

were cultured [12,17].

GST Pull-down AssayGST and GST-fused proHB-EGF-CTF, GST- transforming

growth factor-a (TGF-a)-CTF, GST- amphiregulin (AR)-CTF,

and GST- epiregulin (EPR)-CTF were prepared [12]. After

binding GST and GST-EGFR ligands-CTF to the glutathione

Sepharose beads, lysates containing various FLAG-tagged PLZF

derivatives were incubated with 20 mL of beads for 2 h at 4uC.After washing, bound proteins were analyzed with immunoblot-

ting using an anti-FLAG antibody.

Surface Plasmon Resonance (SPR) SpectroscopyA BIA core S51 SPR systemH (GE Health care) was used to

qualitatively and quantitatively analyze the interactions between

the four immobilized EGFR-ligand-CTFs (i.e. HB-EGF-CTF,

TGF-a-CTF, AR-CTF, and EPR-CTF) and GST-Zn 5-8 of

PLZF, which were measured in resonance units (RU). The

recombinant biotin-EGFR- ligand-CTFs (Peptide Institute, Inc.,

Osaka, Japan) were individually immobilized (,1000 RU)

through amino group coupling on streptavidin (SA) certified

chips. Various concentration of GST-Zn5-8 were added into

running HEPES buffer (10 mM HEPES, pH 7.4, 150 mM NaCl,

0.005% Surfactant P20) at a flow rate of 30 mL/min for 2 min.

The sensor chips were dissociated by a 30-sec injection of 10 mM

Glycine-HCl, pH 1.5. The sensorgrams were analyzed with BIA

evaluation Software (version 4.1; BIAcore). The equilibrium

dissociation constants (’binding constant’; KD) of each reaction

were calculated by dividing the dissociation rates (’off rate’; kd) by

the association rates (’on rate’; ka) (i.e. Kd= kd/ka).

Imaging of CFP Fusion Proteins and Quantitation ofCellular Fractions with Nuclear Localized CFP-PLZFTransiently transfected HT1080 cells, and stable HT1080/HB-

EGF, HT1080/TGF-a, HT1080/AR, and HT1080/EPR cells

were cultured for 24 h and then cells were pretreated with 10 mMKB-R7785 in serum-free medium for 30 min. Cells were then

incubated in serum-free medium with 100 nM TPA for 1 h.

Subcellular localization of CFP fusion protein was examined under

an epifluorescence microscope (Eclipse TE3000, Nikon, Tokyo,

Japan) [12].

Alphascreen AssaysAn Alphascreen systemH (PerkinElmer, Shelton, CT, USA) was

used to screen a library of 9000 chemical compounds (Carna

Bioscience Inc., Kobe, Japan) cyclopaedically for their efficacy in

blocking the interaction between biotinylated-EGFR ligands-CTF

and GST-Zn5-8 of PLZF. 5 ml of 1.25 mg/mL recombinant GST-

ZnF5-8 was incubated with 2.5 mL of 1.25 mg/mL biotin-EGFR

ligands-CTF for 30 min, and 2.5 mL of anti-GST antibody

(1.8 mg/mL) was added for over 1 h. Then 5 mL streptavidin-

coated donor and anti-GST IgG conjugated acceptor beads (1:1

mixture) were incubated for 2 h in the dark. Alphascreen signals

(counts per second) were analyzed with an EnVision microplate

reader (Perkin-Elmer). Assays were conducted at 23uC in buffer

containing 25 mM HEPES, 1 mM MgCl2, 20 mM NaCl, pH 7.4,

0.1% BSA.

Cell Proliferation Assay (CCK-8 Kit Assay)A Cell Counting Kit8 (CCK-8) (Dojindo Laboratories,

Kumamoto, Japan) was used for determining cell proliferation.

Briefly, 26105 cells per well were cultured in 96-well culture plates

with 10% FBS for 24 h. Then, FBS was replaced with serum free

medium for 72 h with or without 30 mM of the 12 candidate

compounds and telmisartan or candesartan (0.3, 3 and 30 mM),

and 30 mM of GW9662 during stimulation with 100 nM of TPA.

10 mL of CCK-8 solution was then added to each well, and

incubated at 37uC for an additional 2 h. Optical density (OD) was

examined at a wave-length of 450 nm.

Monitoring Single Cell GrowthsCells were seeded at a density of 56102 cells per 10 cm of a dish

on day 0 in 10% FBS normal growth medium. After 24 h (on

day1), fairly isolated and morphologically healthy cells were

marked with phase-contrast microscopy. The medium was

replaced with 5% FBS medium (control) with or without

100 nM of TPA, 10 mM of KB-R7785, 100 nM of AG1478,

30 mM of telmisartan and candesartan, and 50 ng/mL of

recombinant HB-EGF (Sigma). The medium was exchanged

transfections, and at least 200 cells expressing CFP-PLZF were examined in each experiment. *P,0.05 for the stimulus effect during TPA treatment vs.no treatment, and **P,0.05 for the inhibitory effect during KB-R7785 treatment vs. TPA treatment. (C) A schematic of the high-throughputAlphascreen system. Upon excitation at 680 nm, ambient oxygen is converted to singlet oxygen (1O2) by a photosensitizer present in the donorbeads. If the acceptor beads are in close proximity (,200 nm), 1O2 transfers its energy to thioxene derivatives present in the acceptor beads leadingto emission of light at 520–620 nm. One complex consists of streptavidin-coated donor beads and biotinylated EGFR ligands-CTF. Another consists ofanti-GST antibody-conjugated acceptor beads and GST-ZF5-8. If the association between EGFR ligand-CTF and ZnF5-8 occurs, the beads are closeenough to allow detection of a signal. Any inhibitors of this interaction would increase the distance between the beads, and the signal would be lost.(D) Alphascreen signals of GST or GST-Zn5-8 incubated with various concentrations of biotin-HB-EGF-CTF. (E) Alphascreen signals of biotin-HB-EGF-CTF incubated with various concentrations of GST or GST-Zn5-8. (E) The binding abilities of HB-EGF, TGF-a, amphiregulin (AR), and epiregulin (EPR), toPLZF with Alphascreen system.doi:10.1371/journal.pone.0056770.g003

Table 1. The direct interaction between Amphiregulin(AR),Epiregulin(EPR)-CTF and PLZF using surface plasmonresonance(SPR) spectroscopy.

ka(M21s21) kd(s21) KD(M)

AR-CTF 1.876104 1.4461023 7.6561028

EPR-CTF 5.616103 8.0661024 1.4661027

A direct interaction between EGFR-ligand-CTF and PLZF (GST-tagged-ZnF5-8)with the SPR system. Rate constants (ka) and dissociation constants (kd) forGST-ZnF5-8 binding to biotin-AR-CTF or biotin-EPR-CTF are performed. KD = kd/ka.doi:10.1371/journal.pone.0056770.t001



every 48 h with fresh medium. Cell morphology and counts were

scored every 24 h [18].

Immunofluorescence MicroscopyCells that formed a colony were switched to serum-free medium

for 48 h and then incubated for 60 min in 5%FBS conditioned

media with or without 10 mM of KB-R7785, 100 nM of AG1478,

50 ng/mL of recombinant HB-EGF, 30 mM of telmisartan, and

30 mM of candesartan. Subsequently, cells were treated with

100 nM of TPA for 90 min, and then fixed with ethanol and

acetone. The subcellular localization of HB-EGF-CTF and PLZF

was analyzed with immunofluorescence. Primary antibodies

against HB-EGF-CTF or PLZF were used. The secondary

antibodies were Alexa Fluor 594 anti-rabbit IgG or Alexa Fluor

anti-mouse IgG (Invitrogen). Images were obtained on an Eclipse

80i fluorescence microscope (Nikon).

Immunoprecipitation and Western BlottingCells in a subconfluent state were switched to serum-free

medium for 48 h and treated with TPA at the indicated times.

Telmisartan and candesartan were added to the cells for 60 min,

prior to the stimuli. Cells were lysed with lysis buffer, and

supernatants were collected and incubated with 1 mg of human

anti-EGFR or anti-HB-EGF-CTF polyclonal antibodies for 2 h at

4uC with end-over-end rotation. Immunoprecipitation and West-

ern blotting were performed [11]. The primary antibodies used

were 4G10 or anti-PLZF, and anti-HB-EGF-CTF. The secondary

antibodies used were anti-mouse IgG HRP-linked or anti-rabbit

IgG HRP-linked antibodies (Cell Signaling Technology). The

membranes were developed on an ECLWestern blotting detection

system (Amersham Biosciences, Buckinghamshire, England).

Silencing the Angiotensin II type 1 Receptor (AT1R) andADAM12Cells were transfected with 30 mM of AT1R and scramble

siRNAs (Santa Cruz, Delaware, CA, USA), and with 10 mM of

ADAM12 and scramble siRNAs (Invitrogen) using lipofectamine

reagent (Invitrogen). 72 h after cells were transfected, the

expression of AT1R and ADAM12 (Santa Cruz Biotechnology)

were analyzed by Western blotting.

StatisticsData are expressed as mean6 SEM. The one-way ANOVA

and Turkey-Kramer’s multiple comparison procedure test were

used to determine statistically significant differences (P,0.05).

Results

Dual Signaling Pathways of EGFR Phosphorylation andHB-EGF C-terminal Fragment Nuclear Translocationduring Cell Proliferation (Quoted from Ref No. [19] andModified.)12-O-tetradecanoylphorbor-13-acetate (TPA) induced epider-

mal growth factor receptor (EGFR) transactivation through

a disintegrin and metalloproteinase (ADAM) - cleaved prohe-

parin-binding EGF-like growth factor (proHB-EGF) and regulates

cell proliferation [10]. In parallel, the carboxy-terminal fragments

(CTF) of proHB-EGF (HB-EGF-CTF) translocate into the inner

nuclear membrane, subsequently exert on the regulation of cell

proliferation by binding nuclear promyelocytic leukemia zinc

finger (PLZF) protein, a transcriptional repressor, thereby causing

its nuclear export (Fig. 1) [19].

We hypothesized that inhibition of nuclear translocation of HB-

EGF-CTF might lead to a new strategy for prevention cell

proliferation during colon cancer development.

However, there is currently no evidence of potent candidates

that specifically block the nuclear translocation signaling of HB-

EGF-CTF.

The CTF of EGFR ligands interacted with the specificpart of PLZF (ZnF5-8)We used the HT1080 cell lines because

Figure 4. The potent candidates as inhibitors and their inhibitory effects on cell proliferation. (A) Twelve candidate compound’sstructural formula based on efficacy for blocking binding of HB-EGF-CTF or AR-CTF to Zn5-8 of PLZF with Alphascreen system. (B) Inhibitory effects oftwelve candidate compounds on TPA-induced cell proliferation in keratinocytes with a cell proliferation assay. 26104 cells were treated in theconditioned media with or without the twelve candidate compounds during TPA stimulation, and absorbance at 450 nm was determined every 12 htill 60 h. **P,0.05 for the inhibitory effects during 12 candidates treatment vs. TPA treatment. (C, D) The inhibitory effects of the candidates,specifically compound no.8016 and telmisartan, on the binding of AR-CTF to PLZF. Plots with the percent inhibition against various concentrations ofcompound no.8016 and telmisartan presented with the IC50 values observed by Alphascreen system. (E) Inhibitory effects of three ARB candidates onthe binding of AR-CTF to PLZF. Plots with the percent inhibition against various concentrations of each inhibitor and inhibition presented with IC50values.doi:10.1371/journal.pone.0056770.g004

Table 2. High-throughput screening (AlphascreenH) systemfor inhibitors that block the interaction between the PLZF andHB-EGF- or AR-CTF.

Biotin-HB-EGF-CTF Biotin-AR-CTF

100 mM 10 mM 100 mM 10 mM

%Inhibition %Inhibition

No.7620 82.1 22.1 76.3 25.6

No.7632 81.7 12.6 74.9 40.5

No.7701 89.6 41.6 71.5 60.5

No.7702 86.7 34.4 77.9 46.3

No.7768 73 24.1 53.2 32.6

No.7787 82.1 10 82.2 14.9

No.7804 85.7 19.5 79.8 49.9

No.7826 82.3 14.5 82.6 30.7

No.7834 90.7 21.5 77.1 43.6

No.7887 93.9 28.4 90.9 39.5

No.7972 67.5 9.9 70.1 34.2

No.8016 71.4 33.1 75 30.4

Telmisartan * * 66.9 48.5

Camdesartan 78.1 12.6 84.1 23.2

Olmesartan * * 51.8 20.5

Losartan * * 68.4 12.9

Compounds: 9,000 R 12, Z’-factor: 0.7360.091, S/B: 20.6,*not calculated.Twelve candidate and four angiotensin II type 1 receptor blocker (ARB) basedon efficacy for blocking binding of HB-EGF-CTF or AR-CTF to Zn5-8 of PLZF withAlphascreen system. Z’-factor showed to be 0.73 (0.5,Z’,1.0), suggesting thereliable data.doi:10.1371/journal.pone.0056770.t002



Figure 5. Inhibitory effects of no.8016 and four ARBs on cell proliferation. (A-D) Inhibitory effects of compound no.8016 (A) and four ARBs(B) on TPA-induced cell proliferation in keratinocytes with a cell proliferation assay. 26104 cells were treated in conditioned media with or without theno.8016 or four ARBs during TPA stimulation, and absorbance at 450 nm was determined every 12 h till 60 h. (A, C) Inhibition of cell proliferation withno.8016 and telmisartan was verified following a wash at 12 h. (D) The cells were also observed with microscopy. **P,0.05 for the inhibitory effectsduring 12 candidates and four ARBs treatment vs. TPA treatment.doi:10.1371/journal.pone.0056770.g005



HT1080 cells express little endogenous HB-EGF precursors, and

they are helpful to analyze the binding sites between HB-EGF-

CTF and PLZF, and nuclear export of PLZF after overexpressing

HB-EGF precursors and PLZF derivatives [12]. We first certified

the interaction between the cytoplasmic domains of EGFR ligands

and PLZF in HT1080 cells expressing the four proEGFR-ligands

and PLZF before screening the potent candidates. Given that

PLZF coimmunoprecipitates with an antibody against the CTF of

proHB-EGF in HT1080 cells overexpressing proHB-EGF and

FLAG-tagged PLZF [12], a coimmunoprecipitation of PLZF was

performed with antibodies against the CTF of the four EGFR

ligands (HB-EGF, TGF-a, AR, EPR), in HT1080 cells. The

FLAG-tagged full length PLZF gene was transiently transfected

into cells stably expressing proHB-EGF, proTGF-a, proAR or

proEPR (Fig. 2A). PLZF was expressed in these cells (Fig. 2B, lane

1). After TPA stimulation for 1 h, FLAG-tagged PLZF protein

which immunoprecipitated with each antibody was detected by

immunoblotting with an anti-FLAG antibody (Fig.2B, lane 2).

Next, we determined the binding regions of PLZF to the

cytoplasmic domains of pro-EGFR ligands in HT1080 in the GST

pull-down assay. A GST pull-down assay showed the interaction

between the EGFR ligand-CTFs and ZnF 5-8 of PLZF. Each

recombinant GST-fused CTF (GST-CTF) pulled down FLAG-

tagged PLZF equally in cell lysates prepared from HT1080 cells

expressing FLAG-tagged PLZF (Fig. 2C, panel 1). We further

investigated the binding regions of the CTFs in PLZF. Various

FLAG-tagged PLZF derivatives (Fig. 2D) were prepared and

incubated with recombinant GST-EGFR ligand-CTFs. The whole

zinc finger region (ZnF1-9) and ZnF5-8 were pulled down equally

by GST-TGF-a-CTF, GST-AR-CTF and GST-EPR-CTF as well

as HB-EGF-CTF, but not GST alone (Fig. 2C, panels 3 and 4).

None of the four GST-CTFs pulled down the deletion mutant

lacking a zinc finger region (BTB+Center) (Fig. 2C, panel 2). We

then tested whether the deletion mutants of ZnF6-7 (PLZF/

DZnF6-7) and ZnF5-8 (PLZF/DZnF5-8) bind the GST-CTFs.

The deletion of ZnF6-7 abrogated the binding of PLZF to GST-

TGF-a-CTF and GST-HB-EGF-CTF, suggesting that TGF-a-CTF and HB-EGF-CTF interact with PLZF via ZnF6-7. In

contrast, PLZF/DZF6-7 bound weakly to GST-AR-CTF and

GST-EPR-CTF (Fig. 2C, panel 5). Lastly, we tested whether ZF5-

8 is a critical region for interacting with AR-CTF or EPR-CTF.

PLZF/DZnF5-8 bound none of the GST-CTFs (Fig. 2C, panel 6).

These data suggest that the ZnF5-8 region is critical for the

interactions between PLZF and the CTFs.

Weak Binding Affinity of HB-EGF-CTF and PLZF wasImportant for the Nuclear Export of PLZFTo further clarify whether the four CTFs directly interact with

PLZF, we analyzed the binding between the CTFs and

recombinant GST-tagged ZnF5-8 (GST-ZnF5-8) using a SPR

system. GST-ZnF5-8 bound with immobilized biotin-AR-CTF

and biotin–EPR-CTF (Fig. 3A) at a concentration of 0.03–

1.0 mM. The KD for the interaction between GST-ZnF5-8 and

biotin-AR-CTF was 76.5 nM, and for GST-ZnF5-8 and biotin-

EPR-CTF was 146 nM (Table 1). In contrast, the KD value for the

interaction between GST-ZnF5-8 and biotin-HB-EGF-CTF was

not calculated due to their rapid dissociation. The binding signal

of GST-ZnF5-8 to immobilized biotin-TGF-a-CTF was observed,

but it was lower than that of biotin-AR-CTF or biotin-EPR-CTF.

Thus, the dissociation constant could not be calculated because of

the absence of dose-dependent binding of GST-ZF5-8 at

a concentration of #0.5 mM and interference with the free SH

residues of the seven cysteines on TGF-a-CTF (Fig. 3A).

Next, to clarify nuclear export of PLZF triggered by the TPA-

inducible shedding of EGFR ligands, the subcellular localization of

the PLZF protein was examined during the intracellular trans-

location of the four CTFs. The expression vector encoding CFP-

PLZF was transfected into HT1080 cells and the stably transfected

HT1080 cells expressing either proHB-EGF, proTGF-a, proARor proEPR. It is known that wild-type HT1080 cells constitutively

express less HB-EGF and thus, the nuclear translocation of HB-

EGF-CTF is not observed [14]. CFP-PLZF was predominantly

localized in the nucleus of HT1080 cells and in the four

transfectants. TPA treatment for 1 h induced a distribution of

CFP-PLZF throughout the entire cytoplasm of HT1080/HB-

EGF, HT1080/TGF-a, and HT1080/EPR cells. However, TPA

did not alter the subcellular localization of CFP-PLZF in HT1080

and HT1080/AR cells. The ratios of CFP-PLZF localizing in the

nucleus after TPA-inducible shedding were significantly low in

HT1080/HB-EGF, HT1080/TGF-a and HT1080/EPR cells

(Fig. 3B). Pre-treatment with 10 mM of KB-R7785, a metallopro-

teinase inhibitor, effectively abrogated the frequency of CFP-PLZF

nuclear export after TPA stimulation in HT1080/HB-EGF,

HT1080/TGF-a and HT1080/EPR cells (Fig. 3B). Intriguingly,

the frequency of the nuclear export of PLZF is inversely correlated

with its affinity. These results indicate that the affinity of CTF-

PLZF interaction affects intracellular localization and function of

PLZF.

High-throughput Screening Assay System for theInhibitors that Block the Interaction between the ZnF5-8of PLZF and AR- or HB-EGF-CTF Narrowed 12 Candidatesand Four ARBsA high-throughput screening assay was developed on an

AlphascreenH to screen for inhibitor that block the interaction

between ZnF5-8 and CTFs. The Alphascreen assay design relies

on the fact that a signal can be detected only when streptavidin-

coated donor and anti-GST antibody-conjugated acceptor beads

are closed within a distance of 200 nm. The beads are brought

into close proximity for this reaction via specific interactions of

ZnF5-8 and CTFs complexes coupled to them (Fig. 3C). In-

creasing concentrations of biotin-HB-EGF-CTF dose-dependently

Figure 6. Cell proliferation through EGFR pathway and nuclear translocation of HB-EGF-CTF during TPA stimulation. TPA-induced cellproliferation through EGFR and nuclear translocation of HB-EGF-CTF signaling. (A) Growth curve assay. HT29 cell numbers which were counted daily,24 h (i.e.day1) after cells were seeded in three dependent colonies that were cultured in conditioned media. The values are means of threeindependent experiments. (B) Cell numbers of colonies cultured in 5% FBS conditioned media with or without TPA, KB-R7785, AG1478, andrecombinant HB-EGF on day6. The cells were also observed with microscopy (6200) *P ,0.05 for the stimulus effect, and **P,0.05 for the inhibitoryeffect. (C) Effects of KB-R7785 and AG1478 on TPA-induced nuclear translocation of HB-EGF-CTF and nuclear export of PLZF. Cells were treated withTPA following preincubation with or without KB-R7785 and AG1478. Immunofluorescent staining with anti-HB-EGF-CTF antibodies (red signals), anti-PLZF antibodies (green) and DAPI (blue), which stains for nuclei was performed. Images were obtained on a fluorescence microscope (6200). Thewhite bar indicated 10 mm. (D) Effects of KB-R7785 on the association between HB-EGF-CTF and PLZF after TPA stimulation. Cells were treated withTPA at various times following preincubation with or without KB-R7785. Blotted samples were probed with antibodies against PLZF afterimmunoprecipitation with anti-HB-EGF-CTF antibody (upper panel). The total amount of HB-EGF-CTF in the immunoprecipitates was determined byreprobing the same blot with an anti-HB-EGF antibody (lower panel).doi:10.1371/journal.pone.0056770.g006



increased the Alphascreen signal, when incubated with GST-

ZnF5-8 (Fig. 3D). Increasing concentrations of GST-ZnF5-8 also

dose-dependently increased the signal when incubated with biotin-

HB-EGF-CTF (Fig. 3E). The Alphascreen signals for the

interaction between biotin-AR-CTF, biotin-EPR-CTF, biotin-

TGF-a-CTF, and biotin-HB-EGF-CTF with GST-ZnF5-8 were

comparable to the binding affinities determined with SPR system

(Fig. 3F).

An analysis with biotin-AR-CTF and GST-Zn5-8 of PLZF

could clearly indicate inhibitory effects of obtained candidates on

interaction of biotin-AR-CTF with GST-Zn5-8 of PLZF because

Alphascreen signal of biotin-AR-CTF was higher than biotin-HB-

EGF-CTF. Based on their efficacy of blocking the binding of AR-

CTF as well as HB-EGF-CTF and to Zn5-8 of PLZF, twelve

candidates, including no.8016, 7701 and 7804 were obtained by

screening 9000 chemical compounds (Fig. 4A). The % inhibition

of 10 mM of no. 8016, 7701 and 7804 on the interaction between

biotin-AR-CTF and GST-Zn5-8 of PLZF were 30.4, 60.5 and

49.9, respectively. The inhibitory effect of compound no.8016 on

the interaction was not the highest (Table 2). However, No.8016

inhibited cell proliferation strongest among twelve candidates in

cell proliferation assay (Fig. 4B). Based on the results of cell

proliferation, we selected no. 8016 as a candidate. The half

maximal inhibitory concentration (IC50) obtained with compound

no. 8016 was 29.9 mM (Fig. 4C).

We then focused on the specific structural formula of the

biphenyl tetrazole in the candidate compound, and further

screened chemical compounds, including ARBs (i.e. telmisartan,

candesartan, olmesartan, and losartan). Based on the efficacy of

blocking the binding of AR-CTF to PLZF, the IC50 of telmisartan,

candesartan, olmesartan, and losartan were 27.8, 21.9, 87.7 and

28.4 mM, respectively, and their chemical structural formula are

shown (Fig. 4D and E). These findings suggest that telmisartan and

candesartan are satisfactory candidates in blocking both HB-EGF-

CTF and AR-CTF from interacting with PLZF.

No.8016 of Twelve Candidate Compounds andTelmisartan Inhibited Cell ProliferationKeratinocyte cells have a lot of endogenous HB-EGF pre-

cursors, and they are helpful to analyze cell proliferation through

HB-EGF-CTF signaling. We used the HT1080 cell lines when

inhibitors were screened [17]. To examine whether the candidates

compounds that blocked the interaction between the CTFs and

PLZF also inhibited cell proliferation, we tested the inhibitory

effects of these twelve candidates and four angiotensin II type 1

receptor (AT1R) blockers (ARBs) on TPA-induced cell pro-

liferation of keratinocyte with a cell proliferation assay. Cells were

treated in conditioned media with or without 30 mM of the twelve

candidate compounds and 30 mM of ARBs for up to 60 h in the

presence of TPA, and the absorbance was measured every 12 h.

Compound no.8016 and telmisartan were found to inhibit cell

proliferation (Fig. 4B, 5B and D). Furthermore, absorbance values

gradually recovered those of TPA-induced levels (i.e. without

compound no.8016 and telmisartan), when cells were washed 12 h

after incubation with compound no.8016 and telmisartan (Fig. 5A

and C).

TPA-induced Cell Proliferation was Regulated throughDual Signaling PathwaysWe next verified that EGFR phosphorylation and nuclear

translocation of HB-EGF-CTF signaling promoted cell prolifera-

tion. TPA is one of the activators of protein kinase C (PKC) d and

induces the ectodomain cleavage of proHB-EGF via ADAM [10].

When HT29 cells were continuously cultured in the presence or

absence of TPA with or without KB-R7785, EGFR tyrosine kinase

inhibitor AG1478, and recombinant HB-EGF (Fig. 6A), there was

a 2-fold increase in cell counts with TPA compared with control

(Fig. 6B, lane 2). To investigate whether ADAM mediated the

TPA-promoted cell proliferation, we tested the effects of KB-

R7785 on TPA-induced cell proliferation. KB-R7785 was used to

block the shedding of the EGFR ligands and subsequent nuclear

translocation of EGFR ligand-CTF. The TPA-induced increase in

cell counts was completely blocked by KB-R7785, where the cells

count was similar to the control levels seen on day 6 (Fig. 6B, lane

3). To investigate whether EGFR activation mediated the TPA-

induced cell proliferation, we tested the effects of AG1478 on

TPA-induced cell proliferation. The increase in cell counts

induced by TPA was completely blocked by AG1478 to control

levels obtained on day 6 (Fig. 6B, lane 4). To investigate whether

nuclear translocation of HB-EGF-CTF mediates TPA-induced cell

proliferation, we also tested the effects of recombinant HB-EGF on

TPA-induced cell proliferation blocked by KB-R7785. KB-R7785

blocked TPA-induced cell proliferation even with EGFR activa-

tion via recombinant HB-EGF (Fig. 6B, lane 5). EGFR activation

with recombinant HB-EGF during inhibition of EGFR ligand

shedding with KB-R7785 did not recover cell proliferation to the

levels achieved with TPA-stimulation. These data suggests that

nuclear translocation of HB-EGF-CTF is the predominant player

involved in cell proliferation.

Then, to examine the effects of KB-R7785 and AG1478 on the

TPA-induced nuclear translocation of HB-EGF-CTF, we in-

vestigated the localization of HB-EGF-CTF and PLZF during

TPA stimulation with or without KB-R7785 or AG1478 in HT29

cells with immunofluorescent staining (Fig. 6C). In control (i.e.

serum-free), HB-EGF-CTF was located in the cell membrane or

cytoplasm, and PLZF was located in the nucleus. Nuclear

translocation of HB-EGF-CTF and nuclear export of PLZF was

evident during TPA stimulation. KB-R7785 blocked the TPA-

induced nuclear translocation of HB-EGF-CTF and nuclear

export of PLZF. However, AG1478 did not inhibit these TPA-

dependent effects.

We also conformed whether TPA induced the association of

HB-EGF-CTF to PLZF. After cultured cells were treated with

TPA in the presence or absence of KB-R7785, the association

between HB-EGF-CTF and PLZF was assessed by immunopre-

cipitation with the anti-HB-EGF-CTF antibody, followed by

Western blotting with the anti-PLZF antibody. The association

between HB-EGF-CTF and PLZF following TPA stimulation

peaked at 120 min. KB-R7785 completely blocked this association

(Fig. 6D).

Telmisartan, but not Candesartan, Inhibited TPA-inducedColon Cancer Cell ProliferationTelmisartan inhibited TPA-induced cell proliferation in a dose-

dependent manner, but not candesartan (Fig. 7A) in a CCK-8 cell

proliferation assay on HT29 cells. And, inhibitory effects of 3 mMand 30 mM with telmisartan on this cell proliferation were also

observed in other colon cancer cell lines (HCT116, SW480, and

CaCo2) (Fig. 7B–D).

In addition, the inhibitory effects of 30 mM of telmisartan and

candesartan on TPA-induced cell proliferation were assessed with

a growth curve analysis and cell proliferation assay. HT29 (Fig. 7E

and F) and HCT116 (Fig. 7G and H) cells were continuously

cultured for up to 5 or 6 days in the presence or absence of TPA

with or without telmisartan and candesartan. The cell counts were

2-fold greater with TPA stimulation compared to control (Fig. 7F

and H, lane 2). These TPA-induced cell counts were inhibited by



telmisartan (Fig. 7F and H, lane 3), but not candesartan (Fig. 7F

and H, lane 4) to approximately control levels. However, in the

absences of TPA stimulation, both telmisartan and candesartan

(Fig. 7F and H, lane 5–6) did not further inhibit cell proliferation

(i.e. below control levels).

It is known that telmisartan has an anti-cancer activity as

a PPARc agonist [20,21]. ADAM12 reportedly mediated TPA-

induced HB-EGF shedding and subsequently nuclear transloca-

tion of HB-EGF-CTF [14]. Next, to clearly investigate the PPARceffect of telmisantan during elimination of TPA-induced cell

proliferation signal by depleting ADAM12, we examined cell

proliferation in the presence or absence of telmisartan during TPA

stimulation with or without ADAM12 depletion (Fig. 7I and J).

ADAM12 depletion inhibited TPA-induced cell proliferation.

Moreover, an additively inhibitory effect of telmisartan was little

observed. In addition, we examined whether the inhibitory effect

of telmisartan on cell proliferation was cancelled by using

GW9662. Telmisartan inhibited TPA-induced cell proliferation.

Moreover, an additively inhibitory effect of telmisartan was little

observed (Fig. 7K).

Nuclear Translocation of HB-EGF-CTF and Nuclear Exportof PLZF, and the Association with HB-EGF-CTF to PLZF,during TPA Stimulation were Blocked by Telmisartan, butnot by CandesartanImmunofluorescent staining was performed in HT29 or

HCT116 cells to determine the intracellular localization of HB-

EGF-CTF and PLZF during TPA stimulation with or without

telmisartan or candesartan. Cultured cells were treated with TPA

after incubation with either telmisartan or candesartan for 1 h.

HB-EGF-CTF was located in the cell membrane or cytoplasm,

and PLZF in the nucleus, in non-stimulated control cells. Nuclear

translocation of HB-EGF-CTF and nuclear export of PLZF were

seen during TPA stimulation. Telmisartan, but not candesartan,

inhibited the TPA-induced nuclear translocation of HB-EGF-CTF

and nuclear export of PLZF (Fig. 8A and B).

We also confirmed whether TPA induced the association

between HB-EGF-CTF and PLZF in HT29 cells. The association

between both was assessed by immunoprecipitation with the anti-

HB-EGF-CTF antibody, followed by Western blotting with the

anti-PLZF antibody, in cultured cells treated with TPA in the

presence or absence of telmisartan or candesartan. The association

peaked at 120 min. Telmisartan, but not candesartan, significantly

blocked this association (Fig. 8C).

The percentage of cells displaying nuclear staining for HB-EGF-

CTF was significantly decreased when HT29 cells were incubated

via TPA with pretreatment of 30 mM of telmisartan, compared to

pretreatment of 30 mM of candesartan (Fig. 8D).

Telmisartan AT1R-independently Inhibited NuclearTranslocation of HB-EGF-CTF and Cell ProliferationTo assess the effect of telmisartan and candesartan on EGFR

phosphorylation, EGFR phosphorylation was assessed by immu-

noprecipitation using an anti-EGFR antibody followed by Western

blotting with an anti-phosphotyrosine antibody. Cells incubated

with or without telmisartan or candesartan were then treated with

TPA at various times. EGFR phosphorylation following stimula-

tion with TPA peaked within 15 min and then gradually decreased

to basal levels within 120 min. EGFR phosphorylation was not

inhibited by either telmisartan or candesartan (Fig. 9A).

Next, both telmisartan and candesartan are antagonists of the

AT1R. And, to determine whether the TPA-induced nuclear

translocation of HB-EGF-CTF and nuclear export of PLZF are

independent of AT1R signaling, the intracellular localization of

HB-EGF-CTF and PLZF was confirmed after knocking down the

AT1R with short interfering RNA (siRNA). First, a complete

knockdown of endogenous AT1R protein was achieved with

siRNA (Fig. 9B). Then, HT29 cells were pre-treated with

telmisartan or candesartan, stimulated with TPA, and immuno-

fluorescent staining with anti-HB-EGF-CTF and anti-PLZF

antibodies was then performed. In control cells, HB-EGF-CTF

was found in the cell membrane or cytoplasm, and PLZF was done

in the nucleus. Nuclear translocation of HB-EGF-CTF and

nuclear export of PLZF were evident during TPA stimulation.

Telmisartan, but not candesartan, inhibited TPA-induced nuclear

translocation of HB-EGF-CTF and nuclear export of PLZF, even

after treatment during AT1R depletion (Fig. 9C).

Finally, we tested the effects of KB-R7785, telmisartan, and

candesartan on TPA-induced cell proliferation after AT1R

depletion with siRNA in HT29 cells with the CCK-8 cell

proliferation assays. The TPA-induced cell proliferation was

inhibited by KB-R7785, and telmisartan, but not by candesartan

in the presence and absence of AT1R depletion (Fig. 9D).

Discussion

We have here shown that TPA promoted cell proliferation

through nuclear translocation signaling of HB-EGF-CTF, and that

telmisartan blocked the binding of HB-EGF-CTF to PLZF, which

in turn, significantly inhibited cell proliferation.

First, we screened for an inhibitor that was capable of blocking

the interaction between HB-EGF-CTF and PLZF. We previously

showed that HB-EGF-CTF was associated with the transcriptional

repressor, PLZF, and triggered its nuclear export [12]. A GST

pull-down assay assessing the in vitro interaction between PLZF

and four CTFs, namely TGF-a-CTF, AR-CTF, EPR-CTF, andHB-EGF-CTF, demonstrated that PLZF interacted with all four

CTFs. However, the deletion mutant, PLZF/DZnF5-8, did not

bind any of the CTFs. These data suggest that the ZnF5-8 region

Figure 7. The inhibitory effects of telmisartan and candesartan on TPA-induced cell proliferation in colon cancer cells. (A-D)Inhibitory effects of telmisartan, but not candesartan, on TPA-induced cell proliferation in HT29 cells (A), HCT116 (B), SW480 (C), and CaCo2 (D) withCCK-8 kit assay. After 24 h of plating, 26105 cells were incubated for 72 h in conditioned media with or without telmisartan, candesartan and/or TPA.Each bar represents the means of six independent experiments. *P,0.05 for the stimulus effect, and **P,0.05 for the inhibitory effect. (E and G) Ingrowth curve assay, HT29 and HCT116 cell numbers were counted daily in three dependent colonies cultured in conditioned media. The values aremeans of three independent experiments.6; Control (5%FBS),&; TPA (100 nM),#; telmisartan (30 mM),N; TPA+telmisartan (30 mM),g; candesartan(30 mM), m; TPA+candesartan (30 mM). (F and H) Cell numbers in colonies cultured in conditioned media with or without TPA, telmisartan andcandesartan on day 5 or 6. *P,0.05 for the stimulus effect, and **P ,0.05 for the inhibitory effect. Lane 1; Control (5% FBS white box), 2; TPA(100 nM), 3; TPA+telmisartan (30 mM), 4; TPA+candesartan (30 mM), 5; telmisartan (30 mM), 6; candesartan (30 mM). (I) Knockdown of ADAM12 withsiRNA. Probing with an anti-ADAM12 antibody (upper panel) and anti- b-actin antibody (lower panel). (J) Inhibitory effects of telmisartan on TPA-induced cell proliferation in HT29 cells with CCK-8 kit assay following knockdown of ADAM12 with siRNA. *P,0.05 for the stimulus effect, and **P,0.05 for the inhibitory effect. (K) Inhibitory effects of telmisartan and GW9662 on TPA-induced cell proliferation in HT29 cells with CCK-8 kit assay.*P,0.05 for the stimulus effect, and **P,0.05 for the inhibitory effect.doi:10.1371/journal.pone.0056770.g007



Figure 8. The effects of telmisartan or candesartan on localization and binding of HB-EGF-CTF and PLZF. (A and B) Immunofluorescentstainings with anti-HB-EGF-CTF antibodies (red), anti-PLZF antibodies (green), and DAPI (blue), in the presence or absence of TPA after pretreatmentwith telmisartan or candesartan. Images were obtained on a fluorescence microscope (6200). The white bar indicated 10 mm. (C) Effects oftelmisartan or candesartan on the TPA-induced association between HB-EGF-CTF and PLZF. HT29 cells were preincubated with or without telmisartanor candesartan. The cells were then treated with TPA for 0, 30, 60 and 120 min. Blotted samples were probed with an anti-PLZF antibody after



is critical for the interactions between PLZF and the CTFs.

Moreover, SPR analysis revealed that the binding affinities of

ZnF5-8 for AR-CTF and EPR-CTF were 76.5 nM and 146 nM,

respectively, which were higher than those of either HB-EGF-CTF

or TGF-a-CTF (Fig. 3A). Immunostaining of the TPA-trigged

PLZF nuclear export demonstrated that PLZF was localized with

in the cytoplasm of HT1080/HB-EGF, HT1080/TGF-a and

HT1080/EPR cells, but not in the HT1080/AR cells (Fig. 3B).

These suggested that AR bound PLZF more strongly than HB-

EGF in the nucleus, but that AR did not feasibly release the

binding in the cytoplasm than HB-EGF. Based on these

observations, the inverse correlation between binding affinity

and nuclear export were evident. Thus, the interaction between

HB-EGF-CTF and PLZF in the nucleus followed by the rapid

release of PLZF from HB-EGF-CTF in the cytoplasm, appears to

regulate its downstream signaling, and was therefore characterized

as a key event during cell proliferation. The SPR system uses

a highly specialized optical technique to analyze biomolecular

interactions and provides both qualitative and quantitative date.

Additionally, in the present study, we established a very useful

assay system to cyclopaedically quantify the interactions between

EGFR ligand-CTFs and ZnF5-8 of PLZF using Alphascreen

(Fig. 3C). Given that the estimations of the interaction between

EGFR ligand-CTFs and ZnF5-8 with Alphascreen were compa-

rable to those obtained with the SPR analysis, Alphascreen was

a useful and powerful for the high-throughput screening of

compounds, which inhibited these interactions with EGFR ligand-

CTFs and its partners, but we need to prepare short peptides with

the specified binding sites between the both and manipulate

binding of those peptides to beads. Thus we ended up using the

Alphascreen methodology, and focused on screening compounds

containing the specific structural formula of biphenyl tetrazole.

This led us to identifying telmisartan and candesartan as potential

candidates.

Subsequently, we attempted to characterize the predominant

signaling pathway involved in the TPA-induced cell proliferation,

specifically EGFR signaling or nuclear translocation of HB-EGF-

CTF in HT29 cells. KB-R7785 was used to block both

intracellular signaling pathway involved in cell proliferation. The

growth curve assay demonstrated that KB-R7785 and AG1478

completely inhibited TPA-induced cell proliferation. Furthermore,

EGFR activation with recombinant HB-EGF during inhibition of

EGFR ligand shedding with KB-R7785 did not recover cell

proliferation to the levels achieved with TPA-stimulation. This

finding suggests that nuclear translocation of HB-EGF-CTF is the

predominant player involved in cell proliferation. Furthermore,

immunofluorescent staining and immunoprecipitation with the

anti-HB-EGF-CTF antibody, followed by Western blotting with

the anti-PLZF antibody, demonstrated that KB-R7785 completely

blocked the nuclear translocation of HB-EGF-CTF, nuclear

export of PLZF and the binding of HB-EGF-CTF to PLZF,

during TPA stimulation. Thus nuclear translocation of HB-EGF-

CTF also plays a central role on TPA-induced cell proliferation.

These observations are consistent with the previous finding that

HB-EGF-CTF on the cell surface translocate to the inner nuclear

membrane [13], full-length forms of HB-EGF did not translocate

to the nucleus in the gut cells overexpressing unshed HB-EGF-

CTF [22], and the suppression of nuclear translocation of HB-

EGF-CTF abrogated cell proliferation in gastric cancer cells [23].

We then tested whether both telmisartan and candesartan

inhibited cell proliferation, nuclear translocation of HB-EGF-

CTF and binding of HB-EGF-CTF to PLZF. Telmisartan, but not

candesartan, significantly inhibited cell proliferation, nuclear

translocation of HB-EGF-CTF and binding of HB-EGF-CTF to

PLZF during TPA stimulation in HT29 and HCT116 cells

(Fig. 7E–H). The differences in the inhibitory effects of telmisartan

and candesartan on the abovementioned cellular function can be

explained by their lipid solubility. Telmisartan is more lipid soluble

than candesartan [24], and thereby, telmisartan can easily pass

through the cell membrane into the cytoplasm and block the

interaction between HB-EGF-CTF and PLZF.

Angiotensin II, a GPCR agonist, also induces EGFR transacti-

vation [14]. In the present study, the TPA-induced nuclear

translocation of HB-EGF-CTF and nuclear export of PLZF were

independent of AT1R signaling during a knockdown of the AT1R

(Fig. 9C). Further, TPA-induced cell proliferation was also

independent of AT1R signaling (Fig. 9D). Based on these findings,

telmisartan inhibited cell proliferation during TPA stimulation by

specifically blocking nuclear translocation of HB-EGF-CTF

AT1R-independently and nuclear transport of PLZF in colon

cancer cells. In addition, there is no evidence of AT1R agonists

that specifically block nuclear translocation of HB-EGF-CTF and

binding of that to PLZF. We here discovered a new property of

telmisartan.

A previous report demonstrated that telmisartan inhibited cell

proliferation through EGFR followed by extracellular-regulated

kinase (ERK) in uterine leiomyoma cells [25]. However, upstream

molecular mechanism of EGFR signaling was unclear. In the

present study, we could elucidate the inhibitory mechanism of

telmisartan. In addition, telmisartan is a partial PPARc agonist to

activate PPARc as the nuclear transcription factor [20]. PPARcagonists induce apoptosis particularly in colon cancer cells and

have inhibitory effects on cell proliferation [26]. Cell proliferation

assays in the presence or absence of telmisartan during TPA

stimulation with ADAM12 depletion and GW9662 showed that

an additively inhibitory effect of telmisartan was little observed. It

has also been demonstrated that PPARc is a nuclear receptor of

prostaglandins and leukotrienes as well, and it down-regulate cell

proliferation in colon cancer cells [26]. These results and fact

suggested that PPARc signaling might not be associated with

ADAM-HB-EGF-CTF signaling, and that inhibitory effect of

telmisartan as PPARc agonist on cell proliferation was weak

during TPA stimulation.

A recent meta-analysis on the incidence of cancers during

treatment with ARBs in15 large clinical trials found that there was

no significant increase in lung, prostate or breast cancer risk and

the overall cancer mortality, in patients receiving ARBs compared

to controls [27]. In addition, telmisartan plus cetuximab inhibited

cell proliferation in a dose-dependent manner, and better than

telmisartan alone, as determined with cell proliferation assay

(unpublished data). Thus, based on these findings, telmisartan plus

combination chemotherapy with anti-EGFR therapy appears to

be a better treatment strategy in patients with colorectal cancer,

immunoprecipitation with the anti-HB-EGF-CTF antibody (upper panel). The total amount of HB-EGF-CTF in the immunoprecipitates was determinedby reprobing the same blot with an anti-HB-EGF-CTF antibody (lower panel). (D)Three staining regions were randomly selected under a6200 fieldand the total for three regions was calculated as the positive rate of cells displaying nuclear staining. The percentage of cells showing nuclear stainingis shown for 100 nM of TPA, and 30 mM of telmisartan and candesartan. Images were obtained on a fluorescence microscope (6200). The white barindicated 10 mm. *P,0.05 for the stimulus effect, and **P,0.05 for the inhibitory effect.doi:10.1371/journal.pone.0056770.g008



Figure 9. The effects of telmisartan and candesartan on HB-EGF-CTF nuclear translocation and cell proliferation during AT1Rdepletion. (A) Effects of telmisartan or candesartan on TPA-induced EGFR phosphorylation. Cells were preincubated with or without telmisartan orcandesartan, and then treated with TPA for 0, 15, 60 and 120 min. Blotted samples were probed with an anti-phosphotyrosine antibody afterimmunoprecipitation with an anti-EGFR antibody (upper panel). The total amount of EGFR in the immunoprecipitates was determined by reprobingthe same blot with an anti-EGFR antibody (lower panel). (B) Effects of telmisartan or candesartan on TPA-induced nuclear translocation of HB-EGF-CTFand nuclear export of PLZF following knockdown of AT1R with siRNA. Probing with an anti-AT1R antibody (upper panel) and anti-b-actin antibody



compared to combination chemotherapy with anti-EGFR therapy

alone.

In conclusion, inhibition of HB-EGF-CTF nuclear translocation

is important for blocking cell proliferation in colon cancer cell.

telmisartan can inhibit cell proliferation by inhibiting the nuclear

translocation of HB-EGF-CTF. Thus, telmisartan and its

derivatives may be potential treatment strategies for the prevention

of cell proliferation and colon cancer development.

Author Contributions

Critical revision of the manuscript for important intellectual content: TJ

SH. Statistical analysis: KO. Study supervision: ST TJ SH. Conceived and

designed the experiments: ST TJ SH KO. Performed the experiments: KO

CM. Analyzed the data: ST TJ SH TM HT TS HK TK KO. Contributed

reagents/materials/analysis tools: YI EN HI SH. Wrote the paper: KO

ST.

References

1. Saika K, Machii R (2011) Time trends in colon, rectum and anus cancer

mortality between 1955 and 2008 in Japan, USA and Europe based on theWHO mortality database. Jpn J Clin Oncol 41: 1153.

2. Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, et al. (2004)

Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractorymetastatic colorectal cancer. N Engl J Med 351: 337–345.

3. Van Cutsem E, Kohne CH, Hitre E, Zaluski J, Chang Chien CR, et al. (2009)Cetuximab and chemotherapy as initial treatment for metastatic colorectal

cancer. N Engl J Med 360: 1408–1417.

4. Amado RG, Wolf M, Peeters M, Van Cutsem E, Siena S, et al. (2008) Wild-typeKRAS is required for panitumumab efficacy in patients with metastatic

colorectal cancer. J Clin Oncol 26: 1626–1634.5. Yotsumoto F, Yagi H, Suzuki SO, Oki E, Tsujioka H, et al. (2008) Validation of

HB-EGF and amphiregulin as targets for human cancer therapy. BiochemBiophys Res Commun 365: 555–561.

6. Higashiyama S, Abraham JA, Miller J, Fiddes JC, Klagsbrun M (1991) A

heparin-binding growth factor secreted by macrophage-like cells that is relatedto EGF. Science 251: 936–939.

7. Yoshida M, Shimura T, Sato M, Ebi M, Nakazawa T, et al. (2012) A novelpredictive strategy by immunohistochemical analysis of four EGFR ligands in

metastatic colorectal cancer treated with anti-EGFR antibodies. J Cancer Res

Clin Oncol.8. Kopp R, Rothbauer E, Ruge M, Arnholdt H, Spranger J, et al. (2003) Clinical

implications of the EGF receptor/ligand system for tumor progression andsurvival in gastrointestinal carcinomas: evidence for new therapeutic options.

Recent Results Cancer Res 162: 115–132.

9. Wang F, Liu R, Lee SW, Sloss CM, Couget J, et al. (2007) Heparin-bindingEGF-like growth factor is an early response gene to chemotherapy and

contributes to chemotherapy resistance. Oncogene 26: 2006–2016.10. Goishi K, Higashiyama S, Klagsbrun M, Nakano N, Umata T, et al. (1995)

Phorbol ester induces the rapid processing of cell surface heparin-binding EGF-like growth factor: conversion from juxtacrine to paracrine growth factor

activity. Mol Biol Cell 6: 967–980.

11. Tanida S, Joh T, Itoh K, Kataoka H, Sasaki M, et al. (2004) The mechanism ofcleavage of EGFR ligands induced by inflammatory cytokines in gastric cancer

cells. Gastroenterology 127: 559–569.12. Nanba D, Mammoto A, Hashimoto K, Higashiyama S (2003) Proteolytic release

of the carboxy-terminal fragment of proHB-EGF causes nuclear export of PLZF.

J Cell Biol 163: 489–502.13. Hieda M, Isokane M, Koizumi M, Higashi C, Tachibana T, et al. (2008)

Membrane-anchored growth factor, HB-EGF, on the cell surface targeted to theinner nuclear membrane. J Cell Biol 180: 763–769.

14. Asakura M, Kitakaze M, Takashima S, Liao Y, Ishikura F, et al. (2002) Cardiachypertrophy is inhibited by antagonism of ADAM12 processing of HB-EGF:

metalloproteinase inhibitors as a new therapy. Nat Med 8: 35–40.

15. Imayama I, Ichiki T, Inanaga K, Ohtsubo H, Fukuyama K, et al. (2006)

Telmisartan downregulates angiotensin II type 1 receptor through activation ofperoxisome proliferator-activated receptor gamma. Cardiovasc Res 72: 184–

190.

16. Yoshida T, Yamagishi S, Nakamura K, Matsui T, Imaizumi T, et al. (2006)Telmisartan inhibits AGE-induced C-reactive protein production through

downregulation of the receptor for AGE via peroxisome proliferator-activatedreceptor-gamma activation. Diabetologia 49: 3094–3099.

17. Hashimoto K, Higashiyama S, Asada H, Hashimura E, Kobayashi T, et al.

(1994) Heparin-binding epidermal growth factor-like growth factor is anautocrine growth factor for human keratinocytes. J Biol Chem 269: 20060–

20066.18. Kinney EL, Tanida S, Rodrigue AA, Johnson JK, Tompkins VS, et al. (2008)

Adenovirus E1A oncoprotein liberates c-Myc activity to promote cell pro-liferation through abating Bin1 expression via an Rb/E2F1-dependent

mechanism. J Cell Physiol 216: 621–631.

19. Tanida S, Kataoka H, Mizoshita T, Shimura T, Kamiya T, et al. (2010)Intranuclear translocation signaling of HB-EGF carboxy-terminal fragment and

mucosal defense through cell proliferation and migration in digestive tracts.Digestion 82: 145–149.

20. Schupp M, Janke J, Clasen R, Unger T, Kintscher U (2004) Angiotensin type 1

receptor blockers induce peroxisome proliferator-activated receptor-gammaactivity. Circulation 109: 2054–2057.

21. Funao K, Matsuyama M, Kawahito Y, Sano H, Chargui J, et al. (2008)Telmisartan is a potent target for prevention and treatment in human prostate

cancer. Oncol Rep 20: 295–300.

22. Shimura T, Yoshida M, Fukuda S, Ebi M, Hirata Y, et al. (2012) Nucleartranslocation of the cytoplasmic domain of HB-EGF induces gastric cancer

invasion. BMC Cancer 12: 205.23. Shimura T, Kataoka H, Ogasawara N, Kubota E, Sasaki M, et al. (2008)

Suppression of proHB-EGF carboxy-terminal fragment nuclear translocation:a new molecular target therapy for gastric cancer. Clin Cancer Res 14: 3956–

3965.

24. Wienen W (2000) A review on telmisartan: A novel, long-acting angiotensin II-receptor antagonist. Cardiovascular Drug Reviews 18: 127–156.

25. Isobe A, Takeda T, Sakata M, Miyake A, Yamamoto T, et al. (2008) Dualrepressive effect of angiotensin II-type 1 receptor blocker telmisartan on

angiotensin II-induced and estradiol-induced uterine leiomyoma cell pro-

liferation. Hum Reprod 23: 440–446.26. Shimada T, Kojima K, Yoshiura K, Hiraishi H, Terano A (2002)

Characteristics of the peroxisome proliferator activated receptor gamma(PPARgamma) ligand induced apoptosis in colon cancer cells. Gut 50: 658–664.

27. Teo KK (2011) Effects of telmisartan, irbesartan, valsartan, candesartan, andlosartan on cancers in 15 trials enrolling 138,769 individuals. J Hypertens 29:

623–635.

(lower panel). (C) Cells were then treated with TPA following preincubation with or without telmisartan or candesartan. Immunofluorescent stainingswith anti-HB-EGF-CTF antibodies (red), anti-PLZF antibodies (green) and DAPI (blue) were performed following knockdown of AT1R with siRNA.Images were obtained on a fluorescence microscope (6400). The white bar indicated 10 mm. (D)The inhibitory effects of telmisartan and candesartanon TPA-induced cell proliferation in HT29 cells with CCK-8 kit assay following knockdown of AT1R with siRNA. *P,0.05 for the stimulus effect, and**P,0.05 for the inhibitory effect.doi:10.1371/journal.pone.0056770.g009



telmisartan inhibits cell proliferation by blocking ...telmisartan inhibits cell proliferation by...

Documents